Comprehensive analysis of anoikis-related long non-coding RNA immune infiltration in patients with bladder cancer and immunotherapy

Abstract

Background: Anoikis is a form of programmed cell death or programmed cell death(PCD) for short. Studies suggest that anoikis involves in the decisive steps of tumor progression and cancer cell metastasis and spread, but what part it plays in bladder cancer remains unclear. We sought to screen for anoikis-correlated long non-coding RNA (lncRNA) so that we can build a risk model to understand its ability to predict bladder cancer prognosis and the immune landscape.

Methods: We screened seven anoikis-related lncRNAs (arlncRNAs) from The Cancer Genome Atlas (TCGA) and designed a risk model. It was validated through ROC curves and clinicopathological correlation analysis, and demonstrated to be an independent factor of prognosis prediction by uni- and multi-COX regression. In the meantime, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis, immune infiltration, and half-maximal inhibitory concentration prediction (IC50) were implemented with the model. Moreover, we divided bladder cancer patients into three subtypes by consensus clustering analysis to further study the differences in prognosis, immune infiltration level, immune checkpoints, and drug susceptibility.

Result: We designed a risk model of seven arlncRNAs, and proved its accuracy using ROC curves. COX regression indicated that the model might be an independent prediction factor of bladder cancer prognosis. KEGG enrichment analysis showed it was enriched in tumors and immune-related pathways among the people at high risk. Immune correlation analysis and drug susceptibility results indicated that it had higher immune infiltration and might have a better immunotherapy efficacy for high-risk groups. Of the three subtypes classified by consensus clustering analysis, cluster 3 revealed a positive prognosis, and cluster 2 showed the highest level of immune infiltration and was sensitive to most chemistries. This is helpful for us to discover more precise immunotherapy for bladder cancer patients.

Conclusion: In a nutshell, we found seven arlncRNAs and built a risk model that can identify different bladder cancer subtypes and predict the prognosis of bladder cancer patients. Immune-related and drug sensitivity researches demonstrate it can provide individual therapeutic schedule with greater precision for bladder cancer patients.

Keywords: anoikis; bioinformatics; bladder cancer; immune status; lncRNA; prognostic model.